Methods and apparatus for correlating biomarkers in non-blood fluids to biomarkers in blood

Abstract

A method for correlating a biomarker in a non-blood body fluid to the same biomarker in the blood of an individual is disclosed, comprising: measuring the biomarker in a non-blood body fluid over a first time period and measuring the same biomarker in the blood of the same individual to establish an R ratio of [NBBF1] / [BB1], where [NBBF1] is the concentration of the biomarker in the non-blood body fluid over the first time period and [BB1] is the concentration of the biomarker in the blood over the first time period; storing the ratio in a memory; measuring the biomarker in the non-blood body fluid over a second time period to determine [NBBF2], where [NBBF2] is the concentration of the biomarker in the non-blood body fluid over the second time period; and correlating the measured [NBBF2] to the R ratio to yield a correlated [BB2] biomarker concentration in the blood of the individual over the second time period. Also disclosed are devices, apparatuses, and methods for correlating glucose concentration in a non-blood body fluid, such as saliva, to glucose in the blood of an individual.

Description

[0001] Cross-references

[0002] This application claims the benefit of U.S. Provisional Patent Application No. 62 / 931,609, filed November 6, 2019, which is incorporated herein in its entirety for all purposes. Background Technology

[0003] This application particularly relates to methods and apparatus for associating biomarkers in non-blood fluids with biomarkers in blood.

[0004] In many cases, measuring biomarkers in non-blood fluids is superior to measuring them in blood because (a) it avoids cutting into the skin or inserting a needle into a vein, and (b) it is easy, rapid, and painless. However, the relationship between the concentration of a biomarker in non-blood fluids and its concentration in blood can vary from person to person, making it difficult and often impossible to use a single universal conversion relationship / constant for conversion across all individuals. For example, the relationship between glucose concentration in saliva and glucose concentration in blood varies from person to person because each person's salivary glands are unique. However, for an individual, the conversion constant / relationship between biomarkers in non-blood fluids and their concentrations in blood is generally fixed. Therefore, it is possible to avoid further blood testing by measuring the conversion constant / relationship between the concentration of biomarkers in non-blood fluids and their concentrations in blood for each individual, and then using that conversion constant / relationship to convert the measurement of biomarkers in non-blood fluids to the measurement of biomarkers in that individual's blood. This is one aspect of the present invention. Summary of the Invention

[0005] A method for associating a biomarker (e.g., glucose) in a non-blood fluid (e.g., saliva) with the same biomarker in an individual's blood.

[0006] The method disclosed in this invention provides a convenient and painless alternative to routine bloodletting procedures for healthcare and medicine.

[0007] In one or more embodiments, the present invention provides:

[0008] A method for associating a biomarker in a non-blood fluid with the same biomarker in an individual's blood, comprising:

[0009] Measure biomarkers in (i) non-blood fluids and (ii) the same biomarkers in the blood of the same individual during a first time period to determine the concentration of biomarkers in non-blood fluids during the first time period [NBBF1] and the concentration of the same biomarkers in the blood during the first time period [BB1].

[0010] Establish a ratio R, where R is the ratio of [NBBF1] to [BB1];

[0011] Store the ratio R in memory (e.g., a non-transient computer-readable medium);

[0012] Measuring biomarkers in non-blood fluids during the second time period to determine the concentration of [NBBF2] in non-blood fluids during the second time period; and

[0013] The measured [NBBF2] was correlated with the ratio R to generate an associated or estimated concentration of the same biomarker in the individual's blood during a second time period.

[0014] Figure 1 An exemplary flowchart of the present invention is provided.

[0015] In some embodiments, the first time period further includes S first time period sub-segments, wherein for each of the S first time period sub-segments, (i) biomarkers in non-blood fluids and (ii) identical biomarkers in the blood of the same individual are measured to determine the concentration of biomarkers in non-blood fluids within the [NBBF1-S] first time period and the concentration of identical biomarkers in the blood within the [BB1-S] first time period, and (b) a ratio R is established, which is the ratio of [NBBF1-s] to [BB1-s], where s represents the "s-th" of the S first time period sub-segments;

[0016] Establish the ratio R of the S first time segment segments, that is, the ratio of [NBBF1-S] to [BB1-S];

[0017] The ratio R' of the first time segment is stored in memory (e.g., a non-transient computer-readable medium);

[0018] Measuring biomarkers in non-blood fluids during the second time period to determine the concentration of [NBBF2] in non-blood fluids during the second time period; and

[0019] The measured [NBBF2] was correlated with the ratio R of the first time period segment to produce the associated or estimated concentration of the same biomarker in the individual's blood during the second time period.

[0020] In some embodiments, the device for measuring biomarkers in non-blood fluids during a first time period is located in the same position as the device for measuring biomarkers in blood. In other embodiments, the two devices are located in different positions.

[0021] In some embodiments, the apparatus for measuring biomarkers in non-blood fluids during a first time period and the apparatus for measuring biomarkers in blood are the same apparatus.

[0022] In some embodiments, the same methods and apparatus of the present invention are used for non-flowing samples (including samples that are difficult to flow on their own).

[0023] The terms “ratio to R”, “R ratio”, and “R” are interchangeable. Attached Figure Description

[0024] If applicable, the accompanying drawings described below are for illustrative purposes only. In some drawings, the figures are to scale, while in others they are not. For clarity, some elements are enlarged when shown in the figures. These drawings are not intended to limit the scope of this disclosure.

[0025] Figure 1 An exemplary flowchart of the present invention. Detailed Implementation

[0026] The following detailed description illustrates certain embodiments of the invention by way of example and not limitation. Section headings and any subheadings used herein are for organizational purposes only and should not be construed as limiting the subject matter in any way. The content under section headings and / or subheadings is not limited to the section headings and / or subheadings, but applies to the entire description of the invention.

[0027] In one or more embodiments, the present invention provides, for example:

[0028] A method for associating a biomarker in a non-blood fluid with the same biomarker in an individual's blood, comprising:

[0029] Measure biomarkers in (i) non-blood fluids and (ii) the same biomarkers in the blood of the same individual during a first time period to determine the concentration of biomarkers in non-blood fluids during the first time period [NBBF1] and the concentration of the same biomarkers in the blood during the first time period [BB1].

[0030] Establish a ratio R, which is the ratio of [NBBF1] to [BB1];

[0031] Store the ratio R in memory (i.e., a non-transient computer-readable medium);

[0032] Measuring biomarkers in non-blood fluids during the second time period to determine the concentration of [NBBF2] in non-blood fluids during the second time period; and

[0033] The measured [NBBF2] was correlated with the ratio R to generate an associated or estimated concentration of the same biomarker in the individual's blood during a second time period.

[0034] An apparatus comprising: a first region for depositing a first sample; and a second region for depositing a second sample, wherein the first and second regions are physically separated and fluidly isolated on the same surface of the same sample card.

[0035] The R-ratio can differ between individuals. For the same individual, the R-ratio can change over a period of time. One aspect of the invention is to establish an R-ratio for an individual and / or to recalibrate their R-ratio after a period of time and / or periodically.

[0036] The method as described in any of the foregoing embodiments further includes periodically calibrating the ratio R to establish a recalibrated [NBBFX] / [BBX] ratio (i.e., RX), where X is the nth time period. In some embodiments, n is 3 to 100. In some other embodiments, n is greater than 100.

[0037] Establish R ratio

[0038] The relevance of the R ratio can depend significantly on factors such as: (i) the number of paired measurements used to establish the R ratio; (ii) the time interval between blood biomarker tests and non-blood fluid biomarker tests in paired tests (because biomarker concentrations in an individual can vary over time); (iii) the instruments and / or methods used to perform the tests (because different instruments and / or methods can give different readings for the same sample); or (iv) any combination thereof.

[0039] The method as described in any of the preceding embodiments, wherein the calibration is selected from: a pair of calibration tests; or several pairs of calibration tests, such as 2 to 50 pairs, and the calibration tests are performed for a period of time to establish the [NBBFA] / [BBA] average ratio, wherein [NBBFA] is the average biomarker concentration in non-blood fluids, and [BBA] is the average biomarker concentration in the blood of an individual during the time period.

[0040] In some embodiments, the method may include one or more of the following:

[0041] Determine R for each individual;

[0042] Recalibrate the R ratio for each individual;

[0043] Establish R-ratios for different times of day;

[0044] Establish an R-ratio for one year;

[0045] Other biomarkers (e.g., metabolites, nucleic acids, cell-free DNA, RNA, etc.);

[0046] Other bodily fluids (e.g., sweat, urine, tears, etc.);

[0047] Using QMax cards capable of measuring blood glucose and salivary glucose to measure analytes in different samples simultaneously or concurrently, i.e., multiplexing substantially different samples simultaneously in the same sample card. For other examples of the applicant’s multiplexing techniques and methods in similar cards or applications, see ESX-103PRV, ESX-106PRV (e.g., pp. 13-17) and ESX-029WO (e.g., pp. 140-146).

[0048] A method for establishing an R ratio includes measuring a pair of BB and NBBF samples at time intervals, including intermediate values ​​or ranges, such as 1 hour or less, 50 minutes or less, 40 minutes or less, 30 minutes or less, 20 minutes or less, 15 minutes or less, 10 minutes or less, and 5 minutes or less.

[0049] A method for establishing an R ratio includes: measuring a pair of BB and NBBF samples, wherein all BB samples are measured using the same instrument, and all NBBF samples are measured by another instrument or by the same method.

[0050] A method for establishing an R ratio includes: measuring BB and NBBF sample pairs, including intermediate values ​​or ranges, within time intervals such as 1 hour or less, 50 minutes or less, 40 minutes or less, 30 minutes or less, 20 minutes or less, 15 minutes or less, 10 minutes or less, and 5 minutes or less; measuring BB and NBBF sample pairs, wherein all BB samples are measured using the same instrument, and all NBBF samples are measured by another instrument or by the same instrument and / or method.

[0051] Correlating NBBF with R ratio to generate relevant BB.

[0052] The correlation of the associated BB generated by correlating NBBF with the R ratio can be significantly dependent on the instrument and / or method used to establish the R ratio and measure NBBF.

[0053] A method for correlating NBBF with an R ratio to generate a correlated BB, comprising: measuring NBBF using the same instruments and / or methods used to establish the R ratio.

[0054] Recalibrate R ratio

[0055] The method as described in any of the foregoing embodiments, wherein periodic calibration is performed over a time period selected, for example, hourly, daily, weekly, monthly, semi-annual, annual, or combinations thereof, including intermediate values ​​and ranges.

[0056] The method as described in any of the foregoing embodiments, wherein the establishment of the recalibrated [NBBFX] / [BBX] ratio is performed within at least one time period selected from the following: every minute, every hour, every morning, every noon, every night, every midnight, every day, every week, every month, every six months, every year, every two years, including intermediate time periods and ranges.

[0057] As described in any of the foregoing embodiments, the association is achieved by solving [BB2] in the following equation:

[0058] [NBBF2] / [BB2] = R,

[0059] Or when rearranging and replacing R:

[0060] [BB2]=[BB1] / [NBBF1][NBBF2].

[0061] As described in any of the foregoing embodiments, the difference between the first time period and the second time period is at least one of the following: from 5 to 10 minutes, from 20 to 30 minutes, from 30 to 40 minutes, from 40 to 50 minutes, from 50 to 60 minutes, per hour, per day, per week, per month, per half year, per year, or per two years, including intermediate values ​​and ranges.

[0062] The method as described in any of the foregoing embodiments, wherein storing the ratio in the memory is implemented using a mobile communication device.

[0063] The method as described in any of the foregoing embodiments, wherein the non-blood fluid is saliva, and the periodic calibration is performed at at least one interval selected from hourly, daily, weekly, monthly, semi-annual, annual, or combinations thereof, including intermediate values ​​and ranges.

[0064] The method as described in any of the foregoing embodiments further includes applying machine learning (ML) to improve the accuracy of the method by manually comparing preliminary, secondary, or tertiary results, for example, generated by currently disclosed devices (cards with separated sample deposition regions) and associated imaging equipment.

[0065] A method for correlating glucose concentration in non-blood fluids with glucose in an individual's blood, comprising:

[0066] Glucose in non-blood fluids and in the blood of the same individual were measured during the first time period to establish the [GNBF1] / [GB1] ratio, where [GNBF1] is the glucose concentration in non-blood fluids during the first time period and [GB1] is the glucose concentration in the individual's blood during the first time period.

[0067] The [GNBF1] / [GB1] ratio is stored in memory;

[0068] Measurements were taken during the second time period of [GNBF2], the concentration of glucose in non-blood fluids; and

[0069] The measured [GNBF2] was correlated with the [GNBF1] / [GB1] ratio to produce a corresponding estimated [BB2] glucose concentration in the individual's blood during the second time period.

[0070] The method as described in any of the foregoing embodiments, wherein the non-blood fluid is saliva.

[0071] The method as described in any of the foregoing embodiments further comprises periodically calibrating the [GNBF1] / [GB1] ratio to establish a recalibrated [GNBFX] / [GBX] ratio, where X is the nth time period, and n is from 3 to 100.

[0072] The method as described in any of the foregoing embodiments, wherein the establishment of the recalibrated [NBBFX] / [BBX] ratio is performed within at least one time interval selected from the following: every minute, every hour, every morning, every noon, every night, every midnight, every day, every week, every month, every six months, every year, every two years, including intermediate time periods and ranges.

[0073] The method as described in any of the foregoing embodiments, wherein storing the [GNBF1] / [GB1] ratio in the memory is performed in the mobile communication device and analyzed by software in the mobile communication device.

[0074] Concentrations in parentheses, such as [NBBF2], are concentrations of biomarkers not found in blood or bodily fluids, and are typically expressed in units such as milligrams per deciliter (mg / dL).

[0075] Non-blood fluids

[0076] Some examples of biomarkers of the present invention include entities such as saliva, sweat, urine, and tears.

[0077] In some embodiments, the same methods and apparatus of the present invention are used for non-flowing samples (including samples that are difficult to flow on their own). One example is the use of QMAX, which is used with two plates to deform a non-flowing sample into a thin layer for measurement.

[0078] Examples of biological samples suitable for use in this invention include, but are not limited to, blood, serum, plasma, nasal swabs, nasopharyngeal washes, saliva, urine, gastric juice, cerebrospinal fluid, tears, feces, mucus, sweat, earwax, oil, glandular secretions, cerebrospinal fluid, tissue, semen, vaginal secretions, interstitial fluid derived from tumor tissue, eye discharge, cerebrospinal fluid, throat swabs, respiration, hair, nails, skin, biopsy, placental fluid, amniotic fluid, umbilical cord blood, lymph, cavity fluid, sputum, pus, microbiota, meconium, breast milk, exhaled condensate, and / or other excretions. Samples may include nasopharyngeal washes. Nasal swabs, throat swabs, fecal samples, hair, nails, earwax, respiration, and other solid, semi-solid, or gaseous samples may be treated in an extraction buffer (e.g., for a fixed or variable time) prior to their analysis. If desired, the extraction buffer or its aliquots may then be treated similarly to other fluid samples. Examples of tissue samples of the object may include, but are not limited to, connective tissue, muscle tissue, nerve tissue, epithelial tissue, cartilage, cancer samples, or bone.

[0079] In some embodiments, the non-flowing substance is first dissolved in solution before being measured.

[0080] biomarkers

[0081] Biomarkers that can be measured using the methods and apparatus of the present invention include molecules, proteins, nucleic acids (e.g., DNA, RNA), cells, or nanoparticles.

[0082] Some examples of biomarkers of the present invention include, but are not limited to:

[0083] Lipomic biomarkers :

[0084] 1) Total cholesterol - a waxy, fatty substance that travels throughout the body in the form of carrier lipoproteins (HDL, LDL and VLDL); a precursor to steroid hormones and bile salts.

[0085] 2) Triglycerides - measure the blood level of the main form of fat stored in our body; also known as triglycerides.

[0086] 3) LDL-C - The amount of cholesterol associated with low-density lipoprotein (LDL) particles in the blood.

[0087] 4) HDL-C - The amount of cholesterol associated with high-density lipoprotein (HDL) particles in the blood.

[0088] Comprehensive metabolomics biomarkers:

[0089] (All include lipid components: total cholesterol, triglycerides, LDL-C, and HDL-C)

[0090] 1) Glucose - blood glucose concentration;

[0091] 2) BUN (blood urea nitrogen) - a marker of renal function.

[0092] 3) Creatine anhydride - a byproduct of creatine decomposition.

[0093] 4) Sodium is a positively charged electrolyte that is essential for muscle contraction, nutrient absorption, nerve function, and pH balance.

[0094] 5) Potassium, a positively charged electrolyte, is essential for muscle contraction, pH balance, nerve signal transduction, and action potentials.

[0095] 6) Chloride - a negatively charged electrolyte; important for maintaining cell homeostasis across the cell membrane and for the production of gastric acid.

[0096] 7) Carbon dioxide (CO2) - measures blood bicarbonate and is an alternative marker for CO2 gas.

[0097] 8) Uric acid - the end product of DNA purine base metabolism and excretion in the kidneys;

[0098] 9) Albumin - the most abundant plasma protein in serum, synthesized in the liver, and binds with other compounds in the blood to form a plasma osmotic gradient.

[0099] 10) Globulins - plasma proteins of different subtypes.

[0100] 11) Calcium plays a variety of roles in the body, including teeth and bone health, blood clotting, neurotransmitter function, muscle contraction, and enzyme activity.

[0101] 12) Phosphorus - a mineral involved in DNA and RNA synthesis, part of ATP, and helps activate enzymes.

[0102] 13) Alkaline phosphatase (ALP) - an enzyme found in the liver, bones, kidneys, small intestine, and placenta.

[0103] 14) Alanine aminotransferase (ALT or SGPT) - an enzyme found in the highest concentration in the liver, but also present in smaller amounts in the heart, muscles and kidneys.

[0104] 15) Aspartate aminotransferase (AST or SGOT) - an intracellular enzyme that is usually elevated due to the destruction of active tissues and cells.

[0105] 16) LDH-lactate dehydrogenase;

[0106] 17) Total bilirubin - a byproduct of red blood cell breakdown and an important component of bile; it functions as an antioxidant.

[0107] 18) GGT - An enzyme mainly found in the liver, kidneys and pancreas.

[0108] 19) Iron – Measured iron bound to transferrin, representing approximately 1 / 3 of the total iron-binding capacity of transferrin. Individually, it is a relatively poor marker of iron status.

[0109] 20)TIBC - Measures the iron-binding capacity of transferrin.

[0110] Other common and recommended biomarkers:

[0111] 1) C-reactive protein

[0112] 2) Cortisol

[0113] 3) DHEA -- Sulfates

[0114] 4) Estimate glomerular filtration rate (eGFR)

[0115] 5) Estradiol

[0116] 6) Ferritin: The body's iron-binding protein.

[0117] 7) Folic acid

[0118] 8) Hemoglobin A1c

[0119] 9) Homocysteine

[0120] 10) Progesterone

[0121] 11) Prostate-specific antigen (PSA)

[0122] 12) Testosterone

[0123] 13) Thyroid-stimulating hormone

[0124] 14) Vitamin D, 25-hydroxy

[0125] 15) Brain natriuretic peptide

[0126] Other blood tests:

[0127] 1) ABO grouping (blood type) - a test used for blood type classification.

[0128] 2) Antinuclear antibody (ANA) - a general test for autoimmune diseases

[0129] 3) Aluminum (blood) - This blood test measures aluminum levels.

[0130] 4) Alcohol

[0131] 5) Arsenic (blood) - This blood test is used to measure arsenic levels.

[0132] 6) Sexually Transmitted Diseases (STDs) - There is no single test for all infectious agents. Screening for STDs typically includes syphilis, trichomoniasis, gonorrhea, chlamydia, herpes, hepatitis, and HIV.

[0133] Complete blood count (CBC) biomarkers :

[0134] RBC: Red blood cell count (red blood cells); Hemoglobin, Hematocrit, MCV: Mean corpuscular volume; MCH: Mean corpuscular hemoglobin; MCHC: Mean corpuscular hemoglobin concentration; Platelets; MPV; RDW: Red blood cell distribution width; Absolute neutrophils; Absolute lymphocytes; Absolute monocytes; Absolute eosinophils; Absolute basophils.

[0135] Anemia biomarkers:

[0136] Erythropoietin (EPO), reticulocytes, ferritin, soluble transferrin receptor (sTrR), folic acid (folate), transferrin, hemoglobin, unsaturated iron-binding capacity (UIBC), iron, vitamin B12, iron saturation.

[0137] Biomarkers for autoimmune diseases:

[0138] Anti-double-stranded DNA Ab (dsDNA), anti-Sjogren's syndrome Ab (SS-B / La), anti-nuclear Ab (ANA), anti-ribonucleoprotein Ab (RNP), anti-smooth muscle Ab (ASMA), anti-mitochondrial Ab (AMA), anti-Sjogren's syndrome Ab (SS-A / Ro), and anti-thyroid peroxidase Ab (TPO).

[0139] Bone disease biomarkers:

[0140] Alkaline phosphatase (ALP), osteocalcin, N-terminal propeptide type 1 collagen (PINP), parathyroid hormone (PTH), bone-specific alkaline phosphatase (BSAP), tartrate-resistant acid phosphatase-5b (TRACP5b), calcium (Ca), vitamin D, 1,25-dihydroxy, C-terminal type 1 collagen telopeptide (CTx), vitamin D, 25-hydroxy, N-terminal type 1 collagen telopeptide (NTx).

[0141] Cardiac markers:

[0142] Apolipoprotein E (Apo E), endothelin-1, brain natriuretic peptide (BNP), LDH, CK, potassium, CKMB, Pro-b-type natriuretic peptide (Pro-BNP), C-reactive protein (CRP), troponin I, and CRPhs (hypersensitive).

[0143] Diabetes biomarkers:

[0144] C-peptide, HbA1c, cholesterol ester transfer protein (CETP), IA-2 antibody, free fatty acids (FFA), insulin, fructosamine, insulin-like growth factor (IGF-1), glucagon, microalbumin, glucose, proinsulin, and glutamate decarboxylase-65 (GAD-65) antibody.

[0145] Endocrine biomarkers:

[0146] Alpha-fetoprotein, growth hormone, adrenocorticotropic hormone (ACTH), growth-releasing factor (GRF), corticosterone, prolactin, cortisol, testosterone, and follicle-stimulating hormone (FSH).

[0147] Gastrointestinal biomarkers:

[0148] Gastrin and lipase.

[0149] Biomarkers of infectious diseases :

[0150] Anti-HBs, Hepatitis B virus load (quantitative), Anti-HBc, Hepatitis C genotype, Anti-HBe, Hepatitis C virus load (qualitative), Anti-HCV, Hepatitis C virus load (quantitative), Anti-HIV / II, HIV virus load (quantitative), Chlamydia trachomatis, Mononucleosis, Cytomegalovirus (CMV), Mycobacterium tuberculosis, HBsAg, Neisseria gonorrhoeae.

[0151] Inflammation / Immune Biomarkers :

[0152] CD cell markers*, immunoglobulins (IgA, IgG, IgM), agglutinin (apolipoprotein J), intracellular adhesion molecule-1 (ICAM-1)-2 complex, C-reactive protein (CRP), IP-10 (IL-12p70)-2 complex, CRPhs (ultrasensitive), ISG15 gene expression, endothelin-1, ISG56 gene expression, fibrinogen, macrophage inflammatory protein (MIP1), glucose-6-phosphate dehydrogenase (G-6-PDH), macrophage inflammatory protein (MIP1), HLA-A, matrix metalloproteinase-2 (MMP-2), HLA-B, matrix metalloproteinase-9 (MMP-9), HLA-C, IFN-induced mononuclear factor The expression of IL-1, IL-10, OAS1 gene, IL-10, OAS2 gene, IL-10(IL-2,IL-4,IFN)-4-complex, OAS3 gene, IL-1, Rantes / CCL5, IL-2(IL-4,IL-10-IFN)-4-complex, serum amyloid (SAA), IL-4(IL-2,IL-10,IFN)-4-complex, tumor growth factor (TGF-1), IL-6, tumor necrosis factor (TNF), IL-8, and vascular adhesion molecule-1 (VCAM-1)-2 complex.

[0153] Lipid metabolism biomarkers:

[0154] Apolipoprotein AI (Apo AI), cholesterol, apolipoprotein AII (Apo AII), HDL-cholesterol, apolipoprotein B-100 (Apo B), LDL-cholesterol (direct), apolipoprotein B48 (Apo B48), lecithin cholesterol acyltransferase (LCAT), apolipoprotein CII (Apo CII), paraoxygenase (PON1), apolipoprotein CIII (Apo CIII), phosphatidylinositol polysaccharide F (PIGF), apolipoprotein E (Apo E), triglycerides.

[0155] Kidney disease biomarkers:

[0156] α-GST, glomerular filtration rate, β-2-microglobulin (serum), microalbumin, β-2-microglobulin (urine), N-acetylglucosidase (NAG), type IV collagen, Pi-GST, creatinine.

[0157] Tumor biomarkers:

[0158] Bcl-2 / IGH, placental growth factor (PlGF), carbohydrate antigen 19-9 (CA19-9), prostate-specific antigen (PSA), carcinogenic embryonic antigen (CEA), vascular endothelial growth factor (VEGF), and fibroblast growth factor (FGFb).

[0159] Thyroid markers:

[0160] Anti-thyroid peroxidase Ab (TPO), thyroid-stimulating hormone (TSH), anti-thyroglobulin Ab, total thyroxine (T4), free thyroxine (FT4), total triiodothyronine (T3), free triiodothyronine (FT3), TSH receptor Ab, and thyroglobulin.

[0161] Pharmacogenomic biomarkers in drug markers:

[0162] HLA-B, ESR (hormone receptor), ERBB2 (HER2), EGFR, ALK, GAA, ERBB2 (HER2), ESR (hormone receptor), PIK3CA, NAT2, NAT2, CYP2D6, CYP2D6, CYP2D6, ESR, PGR (hormone receptor), UGT1A1, CYP2D6, CYP2D6, PML-RARA, G6PD, nonspecific (congenital methemoglobinemia), G6PD, CD274 (PD-L1), gene tag (T-effector), EGFR, ALK, CYP2D6, PDGFRA, F2 (prothrombin) F5 (factor V Leiden), PROC, PROS1, SERPIN1 (antithrombin III), CYP2C9, CD274 (PD-L1), TPMT, NUDT15, UGT1A1, BRAF, UGT1A1, BCR-ABL1 (Philadelphia chromosome), IFNL3 (IL28B), BCR-ABL1 (Philadelphia chromosome), ALK, TNFRSF8 (CD30), CYP2D6, ALK, CYP2C19, CYP2D6, BCR-ABL1 (Philadelphia chromosome), RET, DPYD, MET, HLA-B, HLA-A, NAGS, CYP2D6, CYP2C19, CYP2 D6, G6PD, nonspecific (congenital methemoglobinemia), CYP2C9, ALK, TPP1, EGFR, RAS, CYP2D6, G6PD, nonspecific (congenital methemoglobinemia), G6PD, G6PD, TPMT, CYP2C19, CYP2D6, CYP2C19, CYP2D6, CYP2C19, CYP2D6, BRAF, CYP2D6, HBB, ALK, ROS1, BRAF, G6PD, RAS, IFNL3 (IL28B), EGFR, G6PD, nonspecific (congenital methemoglobinemia), G6PD, CYP2D6, IFNL3 (IL28B), BC R-ABL1 (Philadelphia chromosome), IL2RA (CD25 antigen), nonspecific (genetic susceptibility to malignant hyperthermia), CYP2D6, CYP2D6, CYP2D6, CYP2C19, CYP2D6, CYP2C19, MYCN, ESR, PGR (hormone receptor), UGT1A1, CYP2D6, CYP2D6, CYP2C19, CYP2C9, CYP2C19, CYP2D6, CD274 (PD-L1), chromosome 17p, ACHR, AQP4, CYP2B6, SLCO1B1, IFNL3 (IL28B), CFTR, CYP2D6, GALNS, F5 (factor VLeiden)SERPINC1 (antithrombin III), chromosome 7, chromosome 13, PRF1, RAB27A, SH2D1A, STXBP2, STX11, UNC13D, XIAP (hemophagocytic lymphohistiocytosis), IDH2, BRAF, RAS, NECTIN4, ROS1, NTRK, FGFR, CYP2C9, ERBB2 (HER2), ESR, PGR (hormone receptor), EGFR, G6PD, CYP2D6, CYP2C19, CYP2C19, PROC, PROS1, SERPINC1 (antithrombin III), DMD, ERBB2 (HER2), E SR (hormone receptor), ESR, PGR (hormone receptor), ERBB2 (HER2), CYP2D6, CYP2C9, CYP2C19, CYP2D6, DPYD, DPYD, CYP2D6, CYP2C9, G6PD, CYP2D6, CYP2D6, CYP2C19, HLA-B, ERBB2 (HER2), ESR, PGR (hormone receptor), CYP2D6, EGFR, CYP2D6, CD33, FLT3, CPOX, HMBS, PPOX (acute hepatic porphyria), G6PD, G6PD, G6PD, DMD, ESR, PGR (hormone receptor), nonspecific (N AT), G6PD, chromosome 17p, chromosome 11q, MYD88, CYP2D6, KIT, BCR-ABL1 (Philadelphia chromosome), PDGFRB, FP1L1-PDGFRA, CYP2D6, UGT1A1, AQP4, TTR, BCR-ABL1 (Philadelphia chromosome), HLA-A, microsatellite instability, mismatch repair, CD274 (PD-L1), ALK, EGFR, UGT1A1, chromosome 17p, chromosome 4p;14q, chromosome 14q;16q, nonspecific (genetic susceptibility to malignant hyperthermia), nonspecific (NAT), CYB5R, CYB5R, CFTR, CF TR, CFTR, IDH1, ERBB2 (HER2), ESR, PGR (hormone receptor), CYP2C19, CYP2C19, ERBB2 (HER2), ESR, PGR (hormone receptor), HLA-DQA1, HLA-DRB1, NTRK, IFNL3 (IL28B), chromosome 5q, microsatellite instability, mismatch repair, CYP2C9, ESR, PGR (hormone receptor), nonspecific (congenital methemoglobinemia), G6PD, G6PD, nonspecific (congenital methemoglobinemia), CYP2D6, ALK, ROS1, HBB, F2 (prothrombin), F5 (factor V Leiden), PROC, PROS1, SERPINC1 (antithrombin III), SSTR, G6PDCYP2D6, CYP2C9, G6PD, nonspecific (congenital methemoglobinemia), TPMT, NUDT15, G6PD, CYB5R, G6PD, CYP2D6, CYP2D6, FLT3, NPM1, KIT, GLA, CYP2D6, BCHE, CYP2D6, HPRT1, G6PD, CYP2D6, CYP2D6, ERBB2 (HER2), ESR, PGR (hormone receptor), BCR-ABL1 (Philadelphia chromosome), UGT1A1, BRCA, genomic instability (homologous recombination defect), G6PD, BRAF, CD2 74 (PD-L1), microsatellite instability, mismatch repair, EGFR, ALK, CYP2D6, SMN2, MS4A1 (CD20 antigen), BRCA, ERBB2 (HER2), ESR, PGR (hormone receptor), BRCA, genomic instability (homologous recombination defect), homologous recombination repair, PPP2R2A, PDGFRA, BCR-ABL1 (Philadelphia chromosome), IFNL3 (IL28B), CYP2C19, CYP2D6, EGFR, CYP2C9, CYP2B6, HLA-B, G6PD, nonspecific (congenital methemoglobinemia) ESR (hormone receptor), ERBB2 (HER2), CYP2D6, EGFR, RAS, CYP2C19, CASR, CYP2D6, TTR, UGT1A1, HLA-B, IFNL3 (IL28B), G6PD, BRAF, CD274 (PD-L1), microsatellite instability, mismatch repair, EGFR, ALK, tumor mutation burden, FGFR2, CYP2D6, ERBB2 (HER2), ESR, PGR (hormone receptor), CYP2C9, CYP2C19, HLA-B, CYP2D6, CYP2C9, CYP 2D6, BCR-ABL1 (Philadelphia chromosome), CYP2C19, CYP2C9, CYP3A5, CYP2B6, G6PD, CYB5R, G6PD, nonspecific (NAT), CYP2D6, CYP2D6, CYP2D6, CYP2D6, G6PD, CYP2D6, CYP2C19, ESR (hormone receptor), UGT1A1, EGFR, RAS, G6PD, CYB5R, RAS, ESR, PGR (hormone receptor), ERBB2 (HER2), CYP2C9, CYP2D6, MS4A1 (CD20 antigen), F5 (factor V Leiden), G6PD, nonspecific (congenital methemoglobinemia), SLCO1B1, BRCA, CYP2D6, CYP1A2, BRCA, loss of heterozygosity (homologous recombination defect), UGT1A1, RET, RYR1.IFNL3 (IL28B), CYP2C9, G6PD, nonspecific (congenital methemoglobinemia), ASS1, CPS1, OTC (urea cycle metabolic disorder), IFNL3 (IL28B), IFNL3 (IL28B), IFNL3 (IL28B), G6PD, BCHE, G6PD, G6PD, nonspecific (NAT), G6PD, nonspecific (NAT), TTR, G6PD, BRCA, ERBB2 (HER2), ESR, PGR (hormone receptor), F5 (factor V) Leiden), F2 (prothrombin), CYP2D6, CYP2D6, IFNL3 (IL28B), CYP2D6, TPMT, NUDT15, CYP2D6, CYP2C19, ERBB2 (HER2), RAS, G6PD, G6PD, CYP2D6, ESR (hormone receptor), CYP2D6, BRAF, G6PD, RAS, ERBB2 (HER2), ESR, PGR (hormone receptor), PML-RARA, ACADVL, CPT2, HADHA, HADHB (long chain fatty acid oxidation disorder), CYP2D6, ERBB2 (HER2), CYP2D6, IL12A, IL12B, IL23A, CYP2D6, POLG, nonspecific (urea cycle metabolic disorder), BRAF, RAS, chromosome 17p, chromosome 11q, TP53, IDH1, IDH2, I GH, NPM1, FLT3, CYP2D6, BCR-ABL1 (Philadelphia chromosome), CYP2C19, CYP2D6, HBB, CYP2C9, VKORC1, PROS1, PROC.

[0163] Drug testing markers:

[0164] Alcohol, amphetamines (excluding methamphetamine), methamphetamine, MDMA (ecstasy), barbiturates (excluding phenobarbital), phenobarbital, benzodiazepines, cannabinoids, cocaine, codeine, cotinine (a breakdown product of nicotine), morphine, tricyclic antidepressants (TCAs), LSD, methadone, steroids, PCP.

[0165] Other examples of embodiments are:

[0166] 1. A method for associating a biomarker in a non-blood fluid with the same biomarker in an individual's blood, comprising:

[0167] Biomarkers in non-blood fluids were measured during the first time period, and the same biomarkers in the blood of the same individuals were measured to establish an R ratio equal to [NBBF1] / [BB1], where [NBBF1] is the concentration of the biomarker in non-blood fluids during the first time period, and [BB1] is the concentration of the biomarker in the blood during the first time period.

[0168] Store the R ratio in memory;

[0169] Biomarkers in non-blood fluids were measured during the second time period to determine [NBBF2], where [NBBF2] is the concentration of the biomarker in non-blood fluids during the second time period; and

[0170] The measured [NBBF2] is correlated with the R ratio to generate the associated [BB2] biomarker concentration in the individual's blood during the second time period.

[0171] 2. The method as described in Example 1 further comprises periodically calibrating the ratio R to establish a recalibrated [NBBFX] / [BBX] ratio, where X is the nth time period, and n is from 3 to 100.

[0172] 3. The method as described in Example 2, wherein the calibration is selected from: a pair of calibration tests; or several pairs of calibration tests, and the calibration tests are performed for a period of time to establish the [NBBFA] / [BBA] average ratio, wherein [NBBFA] is the average biomarker concentration in non-blood fluids, and [BBA] is the average biomarker concentration in the blood of an individual during the time period.

[0173] 4. The method as described in Example 2, wherein periodic calibration is performed over a time period selected from hourly, daily, weekly, monthly, semi-annual, annual, or combinations thereof, including intermediate values ​​and ranges.

[0174] 5. The method as described in Example 2, wherein the establishment of the recalibrated [NBBFX] / [BBX] ratio is performed within at least one time period selected from the following: every minute, every hour, every morning, every noon, every night, every midnight, every day, every week, every month, every six months, every year, every two years, including intermediate time periods and ranges.

[0175] 6. The method as described in Example 1, wherein the association is achieved by solving [BB2] in the following formula:

[0176] [NBBF2] / [BB2] = R,

[0177] Or rearrange and replace R:

[0178] [BB2]=[BB1] / [NBBF1][NBBF2].

[0179] 7. The method as described in Example 1, wherein the difference between the first time period and the second time period is at least one of the following: from 5 minutes to 10 minutes, from 20 minutes to 30 minutes, from 30 minutes to 40 minutes, from 40 minutes to 50 minutes, from 50 minutes to 60 minutes, per hour, per day, per week, per month, per half year, per year, or per two years, including intermediate values ​​and ranges.

[0180] 8. The method as described in Example 1, wherein storing the R ratio in the memory is implemented using a mobile communication device.

[0181] 9. The method as described in Example 1, wherein the non-blood fluid is saliva, and the periodic calibration is performed at at least one interval selected from hourly, daily, weekly, monthly, semi-annual, annual, or combinations thereof, including intermediate values ​​and ranges.

[0182] 10. The method of Example 1, further comprising applying machine learning (ML) to improve the accuracy of the method by manually comparing at least one of the following: preliminary results; secondary results; or tertiary results generated by a device having separated sample deposition regions and associated imaging devices or analyte concentration measurement devices.

[0183] 11. A method for correlating glucose concentration in non-blood fluids with glucose in an individual's blood, comprising:

[0184] Glucose in non-blood fluids was measured during the first time period, and glucose in the blood of the same individual was also measured to establish the [GNBF1] / [GB1] ratio (R), where [GNBF1] is the glucose concentration in non-blood fluids during the first time period, and [GB1] is the glucose concentration in the blood of the individual during the first time period.

[0185] The [GNBF1] / [GB1] ratio is stored in memory;

[0186] The glucose concentration in non-blood fluids was measured during the second time period, where [GNBF2] represents the glucose concentration in non-blood fluids during the second time period; and

[0187] The measured [GNBF2] was correlated with the [GNBF1] / [GB1] ratio to produce a corresponding estimated [BB2] glucose concentration in the individual's blood during the second time period.

[0188] 12. The method as described in Example 11, wherein the non-blood fluid is saliva.

[0189] 13. The method as described in Example 11 further comprises periodically calibrating the [GNBF1] / [GB1] ratio to establish a recalibrated [GNBFX] / [GBX] ratio, where X is the nth time period, and n is 3 to 100.

[0190] 14. The method as described in Example 13, wherein the establishment of the recalibrated [GNBFX] / [GBX] ratio is performed within at least one time period selected from the following: every minute, every hour, every morning, every noon, every night, every midnight, every day, every week, every month, every six months, every year, every two years, including intermediate time periods and ranges.

[0191] 15. The method as described in Example 11, wherein storing the [GNBF1] / [GB1] ratio in the memory is performed in a mobile communication device and analyzed by software in the mobile communication device.

[0192] 16. An apparatus comprising:

[0193] The first region for depositing the first sample; and

[0194] The second region is used for depositing the second sample.

[0195] The first and second regions are physically separated and fluidly isolated on the same surface.

[0196] 17. The apparatus as described in Example 16, wherein the first and second regions are spatially separated.

[0197] 18. The apparatus as described in Example 16, wherein the first and second regions are separated by a visible barrier.

[0198] 19. The apparatus as described in Example 16, wherein the first sample is saliva and the second sample is blood.

[0199] 20. The apparatus of Example 16 further includes at least one of the following: a third region for depositing a third sample, a fourth region for depositing a fourth sample, a fifth region for depositing a fifth sample, or a sixth region for depositing a sixth sample.

[0200] Device Example

[0201] Examples of devices used in this invention include, but are not limited to, iMOST, QMAX sample holders, and other devices described in U.S. Patent No. 10,324,009 to Chou and Ding.

[0202] In some embodiments, the sample is formed into a thin layer in which the cells are monolayers that do not significantly overlap with each other.

[0203] In some embodiments, the sample holder has scale markers, image markers, monitoring markers, etc.

[0204] In some embodiments, the sample holder comprises an X-plate having spacers / pillars having a substantially uniform height and a nearly uniform cross-section, the cross-sections being spaced apart from each other by a consistent, defined distance.

[0205] For example, the X-plate is a 175µm thick PMMA column array with a column size of 30×40µm, a column height of 10µm, and a spacing of 80µm, or the iMOST Q card contains an X-plate of 175µm thick PMMA with a column array with a column diameter of 40µm, a column height of 10µm, and a spacing of 120µm.

[0206] In some embodiments, the sample holder utilizes the lateral flow of the sample. The sample is loaded at one location and flows laterally between two plates to another location.

[0207] I.Board

[0208] In this invention, the CROF plate is typically made of any material that (i) can be used with spacers to adjust the thickness of a portion or all of the sample volume, and (ii) has no significant adverse effect on the sample, the measurement, or the objective the plate is intended to achieve. However, in some embodiments, specific materials (and therefore their properties) are used in the plate to achieve certain purposes.

[0209] In some embodiments, the two plates have the same or different parameters for each of the following: plate material, plate thickness, plate shape, plate area, plate flexibility, plate surface properties, and plate optical transparency.

[0210] (i) Plate material. The plate is made of a single material, a composite material, multiple materials, a multilayer material, an alloy, or a combination thereof. Each material used for the plate is an inorganic material, an organic material, or a mixture thereof, with examples of the materials given in the Mat-1 and Mat-2 paragraphs.

[0211] Mat-1: Inorganic materials used in the board include, but are not limited to, glass, quartz, oxides, silicon dioxide, silicon nitride, hafnium oxide (HfO), aluminum oxide (AIO), semiconductors (silicon, GaAs, GaN, etc.), metals (e.g., gold, silver, copper, aluminum, Ti, Ni, etc.), ceramics, or any combination thereof.

[0212] Mat-2: Organic materials used for spacers include, but are not limited to, polymers (e.g., plastics) or amorphous organic materials. Polymer materials used for spacers include, but are not limited to, acrylate polymers, vinyl polymers, olefin polymers, cellulose polymers, non-cellulose polymers, polyester polymers, nylon, cyclic olefin copolymers (COC), poly(methyl methacrylate) (PMMA), polycarbonate (PC), cyclic olefin polymers (COP), liquid crystal polymers (LCP), polyamide (PA), polyethylene (PE), polyimide (PI), polypropylene (PP), polyphenylene ether (PPE), polystyrene (PS), polyoxymethylene (POM), polyetheretherketone (PEEK), polyethersulfone (PES), polyethylene phthalate (PET), polytetrafluoroethylene (PTFE), polyvinyl chloride (PVC), polyvinylidene fluoride (PVDF), polybutylene terephthalate (PBT), fluorinated ethylene propylene (FEP), perfluoroalkoxyalkylene (PFA), polydimethylsiloxane (PDMS), rubber, or any combination thereof.

[0213] In some embodiments, each plate is independently made of at least one of glass, plastic, ceramic, and metal. In some embodiments, each plate independently comprises at least one of glass, plastic, ceramic, and metal.

[0214] In some embodiments, one plate differs from another plate in terms of lateral area, thickness, shape, material, or surface finish. In other embodiments, one plate is identical to another plate in terms of lateral area, thickness, shape, material, or surface finish.

[0215] The material used for the plate is rigid, flexible, or any flexible material in between. Rigidity (e.g., rigid) or flexibility relates to the given pressing force used to bring the plate into the closed configuration.

[0216] In some embodiments, the choice of rigid or flexible plate is determined by the requirement to control the uniformity of sample thickness within a closed configuration.

[0217] In some embodiments, at least one of the two plates is transparent (to light). In some embodiments, at least a portion or several portions of one or both plates are transparent. In some embodiments, the plates are opaque.

[0218] (ii) Plate thickness. In some embodiments, the average thickness of at least one of the plates is 2 nm or less, 10 nm or less, 100 nm or less, 500 nm or less, 1000 nm or less, 2 μm or less, 5 μm or less, 10 μm or less, 20 μm or less, 50 μm or less, 100 μm or less, 150 μm or less, 200 μm or less, 300 μm or less, 500 μm or less, 800 μm or less, 1 mm or less, 2 mm or less, 3 mm or less, or a range between any two values.

[0219] In some embodiments, the average thickness of at least one plate is at most 3 mm, at most 5 mm, at most 10 mm, at most 20 mm, at most 50 mm, at most 100 mm, at most 500 mm, or a range between any two values.

[0220] In some embodiments, the thickness of any one of the plates is not uniform across the entire plate. Using different plate thicknesses at different locations can be used to control plate bending, folding, sample thickness adjustment, etc.

[0221] (iii) Plate shape and area. Generally, the plate can have any shape, as long as it allows for the compression of the sample and adjustment of the sample thickness. However, in some embodiments, a particular shape may be advantageous. The plate shape can be circular, elliptical, rectangular, triangular, polygonal, annular, or any combination of these shapes.

[0222] In some embodiments, the two plates may have the same size or shape, or different sizes or shapes. The area of ​​the plates depends on the application. The area of ​​the plates is at most 1 mm², at most 10 mm², at most 100 mm², at most 1 cm², at most 5 cm², at most 10 cm², at most 100 cm², at most 500 cm², at most 1000 cm², at most 5000 cm², at most 10,000 cm², or more than 10,000 cm², or a range between any two values. The shape of the plates may be rectangular, square, circular, etc.

[0223] In some embodiments, at least one of the plates is in the form of a strip (or band) having a width, thickness, and length. The width is at most 0.1 cm, at most 0.5 cm, at most 1 cm, at most 5 cm, at most 10 cm, at most 50 cm, at most 100 cm, at most 500 cm, at most 1000 cm, or a range between any two values. The length can be the desired length. The strip can be rolled up.

[0224] (iv) Plate surface flatness. In many embodiments, the inner surfaces of these plates are flat or significantly flat and planar. In some embodiments, the two inner surfaces are parallel to each other in the closed configuration. Flat inner surfaces facilitate the quantification and / or control of sample thickness by simply using a predetermined spacer height in the closed configuration. For non-flat inner surfaces of the plates, not only the spacer height but also the exact topology of the inner surface is required to quantify and / or control the sample thickness in the closed configuration. To determine the surface topology, additional measurements and / or corrections are required, which can be complex, time-consuming, and expensive.

[0225] The flatness of the plate surface is relative to the final sample thickness (the final thickness is the thickness under closed construction) and is usually characterized by the term "relative surface flatness," which is the ratio of the change in plate surface flatness to the final sample thickness.

[0226] In some embodiments, the relative surface area is less than 0.01%, 0.1%, less than 0.5%, less than 1%, less than 2%, less than 5%, less than 10%, less than 20%, less than 30%, less than 50%, less than 70%, less than 80%, less than 100%, or a range between any two of these values.

[0227] (v) Plate surface parallelism. In some embodiments, the two surfaces of the plate are significantly parallel to each other. In some embodiments, the two surfaces of the plate are not parallel to each other.

[0228] (vi) Plate Flexibility. In some embodiments, the plates are flexible under compression during the CROF process. In some embodiments, both plates are flexible under compression during the CROF process. In some embodiments, one plate is rigid and the other plate is flexible under compression during the CROF process. In some embodiments, both plates are rigid. In some embodiments, both plates are flexible but have different degrees of flexibility.

[0229] (vii) Optical transparency of the plate. In some embodiments, the plate is optically transparent. In some embodiments, both plates are optically transparent. In some embodiments, one plate is optically transparent while the other is opaque. In some embodiments, both plates are opaque. In some embodiments, both plates are optically transparent but have different degrees of optical transparency. The optical transparency of a plate can refer to a portion or the entire area of ​​the plate.

[0230] (viii) Surface wetting properties. In some embodiments, the plates have an inner surface that wets (e.g., a contact angle less than 90 degrees) the sample, the transfer liquid, or both. In some embodiments, both plates have an inner surface that wets the sample, the transfer liquid, or both; having the same or different wettability. In some embodiments, one plate has an inner surface that wets the sample, the transfer liquid, or both; and the other plate has a non-wetting inner surface (e.g., a contact angle equal to or greater than 90 degrees). Wetting of the inner surface of the plate may refer to a portion or the entire area of ​​the plate.

[0231] In some embodiments, the inner surface of the plate has other nanostructures or microstructures to control the lateral flow of the sample in CROF. Nanostructures or microstructures include, but are not limited to, channels, pumps, etc. Nanostructures and microstructures are also used to control the wetting properties of the inner surface.

[0232] II. Spacer

[0233] (i) Spacer Function. In this invention, the spacers are configured to have one or any combination of the following functions and characteristics: the spacers are configured to (1) control the thickness of the sample or the relevant volume of the sample together with the plates (preferably, the thickness control is precise, uniform, or both over the relevant area); (2) provide the sample with a compressible regulated open flow (CROF) on the plate surface; (3) not occupy a significant surface area (volume) in a given sample area (volume); (4) reduce or increase the sedimentation effect of particles or analytes in the sample; (5) change and / or control the wetting properties of the inner surface of the plates; (6) identify the position, size scale, and / or plate-related information of the plates; or (7) perform any combination of the above.

[0234] (ii) Spacer Structure and Shape. In order to achieve the desired sample thickness reduction and control, in some embodiments, the spacer is fixed to its respective plate. Generally, the spacer can have any shape, as long as the spacer can adjust the sample thickness in the CROF process, but certain shapes are preferred to achieve certain functions, such as better uniformity, less overshoot during pressing, etc.

[0235] These spacers can be a single spacer or multiple spacers (e.g., an array). Some embodiments of multiple spacers are arrays of spacers (e.g., columns) where the spacing distance is periodic or aperiodic, or periodic or aperiodic in certain areas of the plate, or with different distances in different areas of the plate.

[0236] There are two types of spacers: open spacers and closed spacers. Open spacers allow sample flow through them (e.g., the sample flows around and through the spacer; a column acts as a spacer). Closed spacers prevent sample flow (e.g., the sample cannot flow through the spacer; an annular spacer, where the sample is inside the ring). Both types of spacers are used in their height to adjust the final sample thickness in a closed configuration.

[0237] In some embodiments, the spacer is only an open spacer. In some embodiments, the spacer is only a closed spacer. In some embodiments, the spacer is a combination of open and closed spacers.

[0238] The term "cylindrical spacer" means that the spacer has a cylindrical shape, and the cylindrical shape can refer to an object having a height and lateral shape that allows the sample to flow around it during compressed open flow.

[0239] In some embodiments, the lateral shape of the columnar spacer is a shape selected from the group consisting of: (i) circles, ovals, rectangles, triangles, polygons, rings, stars, letter shapes (e.g., L-shapes, C-shapes, letters from A to Z), and number shapes (e.g., shapes such as 0, 1, 2, 3, 4, ... to 9); (ii) shapes from group (i) having at least one rounded corner; (iii) shapes from group (i) having serrated or rough edges; and (iv) any superposition of (i), (ii), and (iii). For multiple spacers, different spacers may have different lateral shapes and dimensions as well as different distances from adjacent spacers.

[0240] In some embodiments, spacers may be and / or may include columns, columns, beads, spheres, and / or other suitable geometries. The lateral shape and size of the spacers (e.g., transverse to the respective plate surface) may be any, except for the following limitations in some embodiments: (i) the spacer geometry does not cause significant errors in measuring sample thickness and volume; or (ii) the spacer geometry will not prevent sample outflow between the plates (e.g., it is not in a closed form). However, in some embodiments, some spacers are required as closed spacers to restrict sample flow.

[0241] In some embodiments, the spacer is shaped with rounded corners. For example, a rectangular spacer has one, several, or all of its rounded corners (like circles rather than 90-degree angles). Rounded corners generally make the spacer easier to manufacture and, in some cases, cause less damage to the biomaterial.

[0242] The sidewalls of the column can be straight, curved, inclined, or of different shapes at different portions. In some embodiments, the spacer is a column with various lateral shapes, sidewalls, and column height-to-lateral area ratios. In a preferred embodiment, the spacer has a column shape designed to allow open flow.

[0243] (iii) Spacer Material. In this invention, the spacers are generally made of any material with a thickness capable of being used together with the two plates to adjust the relevant volume of the sample. In some embodiments, the material used for the spacers differs from the material used for the plates. In some embodiments, the material used for the spacers is at least a portion of the same as the material used for at least one plate.

[0244] The spacers are made of a single material, a composite material, multiple materials, multilayer materials, alloys, or combinations thereof. Each material used for the spacers is an inorganic material, an organic material, or a mixture thereof, with examples of materials given in paragraphs Mat-1 and Mat-2. In a preferred embodiment, the spacers are made of the same material as the plates used in CROF.

[0245] (iv) Mechanical strength and flexibility of the spacer. In some embodiments, the mechanical strength of the spacer is sufficient such that during compression and in the closed configuration of the plate, the height of the spacer is the same as or significantly the same as the height of the plate in the open configuration. In some embodiments, the differences in the spacer between the open and closed configurations can be characterized and predetermined.

[0246] The material used for these spacers can be rigid, flexible, or any flexible material in between. The rigidity is related to the given pressing force used to bring the plate into the closed structure: if the spacer does not deform more than 1% over its height under the pressing force, the spacer material is considered rigid; otherwise, it is flexible. When the spacer is made of a flexible material, the final sample thickness in the closed structure can still be predetermined based on the pressing force and the mechanical properties of the spacer.

[0247] (v) Internal spacers within the sample. To reduce and control the desired sample thickness, particularly to achieve good sample thickness uniformity, in some embodiments, these spacers are placed inside the sample or within the relevant volume of the sample. In some embodiments, one or more spacers with appropriate spacing exist within the sample or its relevant volume. In some embodiments, at least one spacer is inside the sample, at least two spacers are inside the sample or within its relevant volume, or at least "n" spacers are inside the sample or its relevant volume, where "n" can be determined by the sample thickness uniformity or the sample flow characteristics required during CROF.

[0248] (vi) Spacer Height. In some embodiments, all spacers have the same predetermined height. In some embodiments, the spacers have different predetermined heights. In some embodiments, the spacers may be divided into groups or regions, where each group or region has its own spacer height. And in some embodiments, the predetermined height of the spacers is the average height of the spacers. In some embodiments, the spacers have approximately the same height. In some embodiments, a certain percentage of the spacers have the same height.

[0249] The height of the spacer is selected by the desired final sample thickness and remaining sample thickness. The spacer height (predetermined spacer height) and / or sample thickness is 3 nm or less, 10 nm or less, 50 nm or less, 100 nm or less, 200 nm or less, 500 nm or less, 800 nm or less, 1000 nm or less, 1 μm or less, 2 μm or less, 3 μm or less, 5 μm or less, 10 μm or less, 20 μm or less, 30 μm or less, 50 μm or less, 100 μm or less, 150 μm or less, 200 μm or less, 300 μm or less, 500 μm or less, 800 μm or less, 1 mm or less, 2 mm or less, 4 mm or less, or any two of these values.

[0250] The spacer height and / or sample thickness are between 1 nm and 100 nm in one preferred embodiment, between 100 nm and 500 nm in another preferred embodiment, between 500 nm and 1000 nm in a separate preferred embodiment, between 1 μm (e.g., 1000 nm) and 2 μm in another preferred embodiment, between 2 μm and 3 μm in a separate preferred embodiment, between 3 μm and 5 μm in another preferred embodiment, between 5 μm and 10 μm in a separate preferred embodiment, between 10 μm and 50 μm in another preferred embodiment, and between 50 μm and 100 μm in a separate preferred embodiment.

[0251] In some embodiments, the spacer height and / or sample thickness (i) are equal to or slightly larger than the minimum size of the analyte, or (ii) are equal to or slightly larger than the maximum size of the analyte. “Slightly larger” means that it is approximately 1% to 5% larger and is any number between the two values.

[0252] In some embodiments, the spacer height and / or sample thickness are greater than the minimum size of the analyte (e.g., the analyte has an anisotropic shape), but less than the maximum size of the analyte.

[0253] For example, red blood cells have a disk shape with a minimum size of 2 μm (disk thickness) and a maximum size of 11 μm (disk diameter). In one embodiment of the invention, the spacers are selected such that the spacing between the inner surfaces of the plates in the relevant region is 2 μm (equal to the minimum size) in one embodiment, 2.2 μm in another embodiment, or 3 μm (50% larger than the minimum size) in yet another embodiment, but smaller than the maximum size of the red blood cell. This embodiment has certain advantages in blood cell counting. In one embodiment, for red blood cell counting, by setting the inner surface spacing to 2 μm or 3 μm and any number between the two values, the undiluted whole blood sample is confined within this spacing, on average, each red blood cell (RBC) does not overlap with other red blood cells, thus allowing for visually accurate counting of red blood cells. (Excessive overlap between RBCs can lead to serious counting errors).

[0254] In some embodiments of this invention, plates and spacers are used not only to adjust the sample thickness but also, when the plates are in a closed configuration, to adjust the orientation and / or surface density of the analytes / entities in the sample. When the plates are in a closed configuration, a thinner sample thickness results in less analyte / entity per surface area (e.g., a smaller surface concentration).

[0255] (vii) Lateral Dimensions of the Spacer. For open spacers, the lateral dimensions can be characterized by their lateral dimensions (sometimes referred to as width) in the two orthogonal directions, x and y. The lateral dimensions of the spacer may be the same or different in each direction. In some embodiments, the lateral dimension in each direction (x or y) is...

[0256] In some embodiments, the ratio of the lateral dimension in the x-direction to the y-direction is 1, 1.5, 2, 5, 10, 100, 500, 1000, 10,000, or a range between any two values. In some embodiments, different ratios are used to adjust the sample flow direction; the larger the ratio, the flow is along one direction (the larger dimensional direction).

[0257] In some embodiments, different lateral dimensions of the spacers in the x and y directions are used as (a) to use the spacers as scale marks to indicate the orientation of the plate, (b) to use the spacers to generate more sample flow in a preferred direction, or both.

[0258] In a preferred embodiment, the period, width, and height are specified.

[0259] In some embodiments, all spacers have the same shape and size. In other embodiments, each of the spacers has a different lateral dimension.

[0260] For the enclosed spacer, in some embodiments, the internal lateral shape and size are selected based on the total volume of the sample enclosed by the enclosed spacer, wherein the volume dimensions have been described in this disclosure; and in some embodiments, the external shape and size are selected based on the strength required to support the liquid against the spacer and the compressive pressure of pressing the plates.

[0261] (viii) The aspect ratio of the height of the column spacer to the average lateral dimension. In some embodiments, the aspect ratio of the height of the column spacer to the average lateral dimension is 100,000, 10,000, 1,000, 100, 10, 1, 0.1, 0.01, 0.001, 0.0001, 0.00001, or a range between any two values.

[0262] (ix) Spacer Height Accuracy. The spacer height should be precisely controlled. The relative accuracy of the spacers (e.g., the ratio of deviation to the desired spacer height) is less than 0.001%, less than 0.01%, less than 0.1%; less than 0.5%, less than 1%, less than 2%, less than 5%, less than 8%, less than 10%, less than 15%, less than 20%, less than 30%, less than 40%, less than 50%, less than 60%, less than 70%, less than 80%, less than 90%, less than 99.9%, or a range between any two values.

[0263] (x) Spacing distance. The spacers can be a single spacer or multiple spacers in relevant areas of the plate or sample. In some embodiments, the spacers on the plate are configured and / or arranged in an array, and the array is periodic, aperiodic, or periodic in some locations on the plate and aperiodic in others.

[0264] In some embodiments, the periodic array of spacers has a lattice of square, rectangular, triangular, hexagonal, polygonal or any combination thereof, wherein the combination means that different positions of the plate have different spacer lattices.

[0265] In some embodiments, the spacing distance of the spacer array is periodic (e.g., uniform spacing distance) in at least one direction of the array. In some embodiments, the spacing distance is configured to improve the uniformity between the plate spacings in a closed configuration.

[0266] The distance between adjacent spacers (i.e., the spacing distance) is 1 μm or less, 5 μm or less, 10 μm or less, 20 μm or less, 30 μm or less, 40 μm or less, 50 μm or less, 60 μm or less, 70 μm or less, 80 μm or less, 90 μm or less, 100 μm or less, 200 μm or less, 300 μm or less, 400 μm or less, or any two values ​​in between.

[0267] In some embodiments, the spacing is 400 or less, 500 or less, 1 mm or less, 2 mm or less, 3 mm or less, 5 mm or less, 7 mm or less, 10 mm or less, or any range between these values. In some embodiments, the spacing is 10 mm or less, 20 mm or less, 30 mm or less, 50 mm or less, 70 mm or less, 100 mm or less, or any range between these values.

[0268] The distance between adjacent spacers (e.g., the spacing distance) is selected such that, for a given property of the plate and the sample, in a closed configuration of the plate, in some embodiments, the sample thickness variation between two adjacent spacers is at most 0.5%, 1%, 5%, 10%, 20%, 30%, 50%, 80%, or any range between these values; or in some embodiments, at most 80%, 100%, 200%, 400%, or any range between two values.

[0269] Clearly, in order to maintain a given variation in sample thickness between two adjacent spacers, closer spacing is required when using more flexible plates.

[0270] Specify the accuracy of the interval distance.

[0271] In a preferred embodiment, the spacers are a periodic square array, wherein the spacers are columns with a height of 2 to 4 μm, an average lateral dimension of 5 to 20 μm, and a spacing of 1 μm to 100 μm.

[0272] In a preferred embodiment, the spacers are a periodic square array, wherein the spacers are columns with a height of 2 to 4 μm, an average lateral dimension of 5 to 20 μm, and a spacing of 100 μm to 250 μm.

[0273] In a preferred embodiment, the spacers are a periodic square array, wherein the spacers are columns with a height of 4 to 50 μm, an average lateral dimension of 5 to 20 μm, and a spacing of 1 μm to 100 μm.

[0274] In a preferred embodiment, the spacers are a periodic square array, wherein the spacers are columns with a height of 4 to 50 μm, an average lateral dimension of 5 to 20 μm, and a spacing of 100 μm to 250 μm.

[0275] The period of the spacing array is between 1 nm and 100 nm in one preferred embodiment, between 100 nm and 500 nm in another preferred embodiment, between 500 nm and 1000 nm in a separate preferred embodiment, between 1 μm (e.g., 1000 nm) and 2 μm in another preferred embodiment, between 2 μm and 3 μm in a separate preferred embodiment, between 3 μm and 5 μm in another preferred embodiment, between 5 μm and 10 μm in a separate preferred embodiment, and between 10 μm and 50 μm in another preferred embodiment, between 50 μm and 100 μm in a separate preferred embodiment, between 100 μm and 175 μm in a separate preferred embodiment, and between 175 μm and 300 μm in a separate preferred embodiment.

[0276] (xi) Spacer density. Spacers are arranged on the respective plates at a surface density greater than 1 / um², greater than 1 / 10um², greater than 1 / 100um², greater than 1 / 500um², greater than 1 / 1000um², greater than 1 / 5000um², greater than 1 / 0.01mm², greater than 1 / 0.1mm², greater than 1 / 1mm², greater than 1 / 5mm², greater than 1 / 10mm², greater than 1 / 100mm², greater than 1 / 1000mm², greater than 1 / 10000mm², or greater than 1 / 10000mm², or within any two values.

[0277] (3) These spacers are configured not to occupy a significant surface area (volume) in a given sample region (volume);

[0278] (xii) The ratio of spacer volume to sample volume. In many embodiments, the ratio of spacer volume (e.g., the volume of the spacer) to sample volume (e.g., the volume of the sample), and / or the ratio of spacer volume to sample volume within a relevant volume of the sample, is controlled to achieve certain advantages. These advantages include, but are not limited to, uniformity of sample thickness control, uniformity of analytes, and sample flow properties (e.g., flow rate, flow direction, etc.).

[0279] In some embodiments, the ratio of the spacer volume (r) to the sample volume, and / or the ratio of the spacer volume to the sample volume within the relevant volume of the sample, is less than 100%, at most 99%, at most 70%, at most 50%, at most 30%, at most 10%, at most 5%, at most 3%, at most 1%, at most 0.1%, at most 0.01%, at most 0.001%, or any range between these values.

[0280] (xiii) Spacers fixed to the plate. The spacing and orientation of the spacers, which play a key role in this invention, are preferably maintained during the process of bringing the plate from the open structure to the closed structure, and / or preferably predetermined before the process from the open structure to the closed structure.

[0281] In some embodiments of this disclosure, spacers are secured to one of the plates before they are brought into the closed configuration. The term “spacer secured to its corresponding plate” means that the spacer is attached to the plate and remains attached during the use of the plate. An example of “spacer secured to its corresponding plate” is that the spacer is made integrally of a single piece of material from the plate, and the position of the spacer relative to the plate surface does not change. An example of “spacer not secured to its corresponding plate” is that the spacer is bonded to the plate by adhesive, but during the use of the plate, the adhesive cannot hold the spacer in its original position on the plate surface (e.g., the spacer moves away from its original position on the plate surface).

[0282] In some embodiments, at least one of these spacers is fixed to its corresponding plate. In some embodiments, two spacers are fixed to their respective plates. In some embodiments, a majority of the spacers are fixed to their respective plates. In some embodiments, all spacers are fixed to their respective plates.

[0283] In some embodiments, the spacer is integrally fixed to the plate.

[0284] In some embodiments, the spacer is attached to its respective plate by one or any combination of the following methods and / or configurations: attaching, bonding, fusing, embossing, and etching.

[0285] The term "embossing" refers to the process of integrally fixing spacers and plates by embossing (e.g., treading) a sheet of material to form spacers on the surface of a plate. The material can be a single layer or multiple layers.

[0286] The term "etched" refers to the process of fixing spacers and plates together by etching a sheet of material to form spacers on the surface of the plate. The material can be a single layer or multiple layers.

[0287] The term "fusion to" refers to the integral fixation of spacers and plates by attaching them together, with the original materials of the spacers and plates fused together and a clear material boundary existing between the two materials after fusion.

[0288] The term "bonded to" means that the spacer and the plate are fixed together as a whole by adhesive bonding.

[0289] The term "attached to" means that spacers and plates are connected together.

[0290] In some embodiments, the spacer and the plate are made of the same material. In other embodiments, the spacer and the plate are made of different materials. In other embodiments, the spacer and the plate are formed integrally. In yet another embodiment, the spacer has one end fixed to its respective plate, while that end is open to accommodate the different constructions of the two plates.

[0291] In other embodiments, each of the spacers is independently attached to, bonded to, fused to, imprinted to, and etched into at least one of the corresponding plates. The term "independently" means securing a spacer to its corresponding plate by the same or different methods selected from those of attaching, bonding, fusing, imprinting, and etching in the corresponding plate.

[0292] In some embodiments, the distance between at least two spacers is predetermined (“predetermined spacing distance” means that the distance is known when the user uses the board).

[0293] In some embodiments of all the methods and apparatuses described herein, additional spacers exist in addition to the fixed spacers.

[0294] (xiv) Specific sample thickness. In this invention, it has been observed that a larger plate holding force (e.g., the force that holds two plates together) can be obtained by using a smaller plate spacing (for a given sample area) or a larger sample area (for a given plate spacing) or both.

[0295] In some embodiments, at least one of these plates is transparent in the region surrounding the relevant area, each plate has an inner surface configured to contact the sample in the closed configuration; in the closed configuration, the inner surfaces of the plates are substantially parallel to each other; except where there are spacers, the inner surfaces of the plates are substantially planar; or any combination thereof.

[0296] Spacers can be fabricated on a board in various ways using photolithography, etching, embossing (nanoimprinting), deposition, stripping, fusion, or combinations thereof. In some embodiments, the spacers are directly embossed or imprinted onto the board. In some embodiments, the spacers are imprinted into a material (e.g., plastic) deposited on the board. In some embodiments, spacers are fabricated by directly embossing the surface of a CROF board. Nanoimprinting can be performed using a roll-to-roll technique or roll-to-planar nanoimprinting. This method offers significant economic advantages, thus reducing costs.

[0297] In some embodiments, spacers are deposited on a plate. Deposition can be evaporation, bonding, or peeling. In bonding, spacers are first fabricated on a carrier, and then the spacers are transferred from the carrier to the plate. In peeling, removable material is first deposited on the plate, creating pores in the material; the bottom of the pores exposes the plate surface, then spacer material is deposited into the pores, after which the removable material is removed, leaving only the spacers on the plate surface. In some embodiments, the spacers deposited on the plate are fused to the plate. In some embodiments, the spacers and the plate are manufactured in a single process. This single process includes embossing (e.g., embossing, molding) or bonding.

[0298] In some embodiments, at least two of the spacers are attached to the respective plates by different manufacturing methods, and optionally said different manufacturing methods include at least one of deposition, bonding, fusion, imprinting and etching.

[0299] In some embodiments, one or more of the spacers are fixed to the respective plate by manufacturing methods such as bonding, fusion, embossing, or etching, or any combination thereof.

[0300] In some embodiments, manufacturing methods for forming such monolithic spacers on a plate include bonding, fusion, imprinting, or etching, or any combination thereof.

[0301] Machine Learning

[0302] Detailed information about the network is described in various publications, including International Application (IA) No. PCT / US2018 / 017504, filed February 8, 2018, and No. PCT / US2018 / 057877, filed October 26, 2018, which are each incorporated herein by reference for all purposes.

[0303] One aspect of this invention provides a framework for machine learning and deep learning for analyte detection and localization. Machine learning algorithms are those capable of learning from data to detect, segment, and classify analytes from images of samples. A more rigorous definition of machine learning is "a computer program is said to learn from experience E about a class of tasks T and a performance measurement P if its performance at a task T improves with experience E as measured by P." It explores algorithms capable of learning and predicting from data—algorithms that overcome static program instructions by building models based on sample inputs and making data-driven predictions or decisions.

[0304] Deep learning is a specific type of machine learning based on a set of algorithms that attempt to model high-level abstractions in data. In a simple case, there might be two sets of neurons: neurons that receive input signals and neurons that send output signals. When an input layer receives input, it passes a modified version of the input to the next layer. In deep networks, there are many layers between the input and output (and these layers are not made up of neurons, but it's helpful to think of it this way), allowing the algorithm to use multiple processing layers consisting of multiple linear and nonlinear transformations.

[0305] One aspect of this invention is two machine learning-based methods for object detection and localization. The first method is a deep learning method, and the second method is a combination of deep learning and computer vision methods.

[0306] (i) Deep learning approach. In the first approach, the disclosed object detection and localization workflow consists of two phases: training and prediction. We describe the training and prediction phases in the following paragraphs.

[0307] (a) Training phase

[0308] During the training phase, annotated training data is fed into the convolutional neural network (CNN). A CNN is a specialized neural network used to process data, featuring grid-like, feedforward, and hierarchical network topologies. Examples of data include time-series data and image data; time-series data can be considered as a 1D grid sampled at regular time intervals, and image data as a 2D grid of pixels. Convolutional networks have achieved success in practical applications. The name "convolutional neural network" indicates that the network employs a mathematical operation called convolution. Convolution is a specialized linear operation. A CNN is simply a neural network that uses convolution instead of general matrix multiplication in at least one of its layers.

[0309] The machine learning model receives one or more images of a sample containing analytes acquired by an imager on a sample-holding QMAX device as training data. The training data is annotated with analytes to be determined, where the annotations indicate whether the analytes are in the training data and their locations in the images. Annotations can be in the form of tight bounding boxes that fully contain the analytes or the center location of the analytes. In the latter case, the center location is further transformed into a Gaussian kernel in a circle or dot plot covering the analyte.

[0310] When training large amounts of data, training machine learning models faces two challenges: annotation (usually done by humans) is time-consuming, and training is computationally expensive. To overcome these challenges, the training data can be divided into small patches, and then these patches, or portions of these patches, can be annotated and trained. The term "machine learning" refers to algorithms, systems, and devices in the field of artificial intelligence that typically use statistical techniques to train artificial neural networks from data without explicit programming.

[0311] Annotated images are fed into a machine learning (ML) training module, where a model trainer trains an ML model based on the training data (annotated sample images). The input data is iteratively fed into the model trainer multiple times until a specific stopping criterion is met. The output of the ML training module is an ML model—a computational model built from data according to the training process in machine learning, which gives the computer the ability to independently perform certain tasks (e.g., detecting and classifying objects).

[0312] During the prediction (or inference) phase, a trained machine learning model is applied by a computer. Examples of machine learning models include ResNet, DenseNet, etc., which are also referred to as "deep learning models" due to the depth of connected layers in their network structure. In some implementations, the Caffe library with fully convolutional networks (FCNs) is used for model training and prediction, and other convolutional neural network architectures and libraries, such as TensorFlow, can also be used.

[0313] The training phase produces a model that will be used in the prediction phase. This model can be reused in the prediction phase to measure the input. Therefore, the computational unit only needs access to the generated model. It does not need access to the training data, nor does it need to run the training phase again on the computational unit.

[0314] (b) Forecasting Phase

[0315] During the prediction / inference phase, the detection component is applied to the input image, which is then fed into a prediction (inference) module preloaded with the trained model generated from the training phase. The output of the prediction phase can be a bounding box containing the detected analytes and their center locations, a dot plot indicating the location of each analyte, or a heatmap containing information about the detected analytes.

[0316] When the output of the prediction phase is a list of bounding boxes, the number of analytes in the image of the sample used for measurement is characterized by the number of detected bounding boxes. When the output of the prediction phase is a dot plot, the number of analytes in the image of the sample used for measurement is characterized by the integral of the dot plot. When the output of the prediction phase is a heatmap, the localization component is used to identify locations, and the number of analytes detected from those locations is characterized by entries in the heatmap.

[0317] One implementation of the localization algorithm involves sorting the heatmap values ​​into a one-dimensional ordered list from highest to lowest. The pixel with the highest value is then selected and removed from the list along with its neighbors. This process is repeated to select the pixel with the highest value from the list until all pixels have been removed.

[0318] In the detection component using heatmaps, the input image, along with the model generated from the training phase, is fed into a convolutional neural network, and the output of the detection phase is a pixel-level prediction in the form of a heatmap. The heatmap can have the same size as the input image, or it can be a scaled-down version of the input image, and it serves as input to the localization component. We disclose an algorithm for locating the center of an object based on the heatmap. The main idea is to iteratively detect local peaks based on the heatmap. After a peak is located, we compute a local region surrounding the peak but with a smaller value. We remove this region from the heatmap and find the next peak from the remaining pixels. This process is repeated until all pixels are removed from the heatmap.

[0319] In some embodiments, the present invention provides a localization algorithm to sort heatmap values ​​from highest to lowest into a one-dimensional ordered list. The pixel with the highest value is then selected and removed from the list along with its neighbors. This process is repeated to select pixels with the highest values ​​from the list until all pixels have been removed from the list.

[0320]

[0321]

[0322] After sorting, the heatmap is a one-dimensional ordered list where heatmap values ​​are sorted from highest to lowest. Each heatmap value is associated with its corresponding pixel coordinates. The first item in the heatmap is the item with the highest value, which is the output of the pop (heatmap) function. A disk is created, with the pixel coordinates of the disk containing the highest heatmap value at its center. Then, all heatmap values ​​whose pixel coordinates fall within the disk are removed from the heatmap. This algorithm repeatedly pops the highest value in the current heatmap, removing the disks surrounding it, until all items are removed from the heatmap.

[0323] In an ordered list heatmap, each item knows which items to continue with and which to follow. When removing an item from the ordered list, we make the following changes:

[0324] • Assume the removed project is xr, its subsequent project is xp, and its follow-up project is xf.

[0325] • For project XP, its successor project is redefined as a successor project within a removed project. Therefore, the successor project for XP is now XF.

[0326] • For removing an xr project, undefine its subsequent projects and remove it from the ordered list.

[0327] • For the subsequent project xf, its progress item is redefined as a progress item in the removal project. Therefore, the progress item of xf is now xp.

[0328] The localization algorithm ends after all items are removed from the ordered list. The number of elements in the set locus will be the count of the analytes, and the location information will be the pixel coordinates of each s in the set locus.

[0329] Another embodiment searches for local peaks, which are not necessarily the peaks with the highest heatmap values. To detect each local peak, we start from a random starting point and search for local maxima. After finding a local peak, we calculate the local region surrounding the peak but with smaller values. We remove this region from the heatmap and find the next peak from the remaining pixels. We repeat this process until all pixels are removed from the heatmap.

[0330]

[0331]

[0332] This is a breadth-first search algorithm starting from s, with a modified condition for visiting points: only if heatmap[p] > 0 and heatmap[p] <= heatmap[q] are the neighbors p of the current position q added for coverage. Therefore, each pixel in the cover has a non-descent path to the local peak s.

[0333] (ii) Hybrid approach using deep learning and computer vision methods. In this second approach, detection and localization are performed using computer vision algorithms, while classification is performed using deep learning algorithms. The computer vision algorithms detect and localize potential candidates for the analyte, and the deep learning algorithms classify each potential candidate as a true analyte or a false analyte. The locations of all true analytes (along with the total count of true analytes) are recorded as the output.

[0334] (a) Detection. Computer vision algorithms detect potential candidates based on the characteristics of the analyzed object, including but not limited to intensity, color, size, shape, and distribution. Preprocessing schemes can improve detection. Preprocessing schemes include contrast enhancement, histogram adjustment, color enhancement, noise reduction, smoothing, and defocusing. The preprocessed input image is fed into a detector. The detector informs the detector of the presence of potential candidates for the analyzed object and provides an estimate of their location. Detection can be based on the structure of the analyzed object (e.g., edge detection, line detection, circle detection, etc.), connectivity (e.g., blob detection, connected components, contour detection, etc.), intensity, color, shape, etc., using schemes such as adaptive thresholding.

[0335] (b) Localization. Following detection, computer vision algorithms localize each possible candidate of the analyte by providing its boundary or a tight bounding box containing it. This can be achieved using object segmentation algorithms such as adaptive thresholding, background subtraction, flood filling, mean shift, watershed, etc. Typically, localization is combined with detection to produce the detection results and the location of each possible candidate of the analyte.

[0336] (c) Classification. Deep learning algorithms, such as convolutional neural networks, have achieved state-of-the-art visual classification. We employ deep learning algorithms to classify each possible candidate of the object to be analyzed. Various convolutional neural networks can be used for object classification, such as VGGNet, ResNet, MobileNet, DenseNet, etc.

[0337] Given each possible candidate for an analyte, the deep learning algorithm computes high-level features through layers of neurons via convolutional and nonlinear filters to extract features that distinguish the analyte from the non-analytes. A fully convolutional layer combines these high-level features into the classification result, telling it whether it is a true analyte or the probability of it being a true analyte.

[0338] Furthermore, these two approaches can be further extended and combined, as will be apparent to those skilled in the art. By applying computer vision algorithms only to the preprocessing of images, the hybrid of deep learning and computer vision can become more deep learning-oriented, with each step in detection, localization, and classification based on a dedicated deep learning model or using a single deep learning model (e.g., RetinaNet) for one-step detection and classification.

Claims

1. A method for associating a biomarker in a non-blood fluid with the same biomarker in an individual's blood, comprising: Biomarkers in non-blood fluids were measured during the first time period, and the same biomarkers in the blood of the same individuals were measured to establish an R ratio equal to [NBBF1] / [BB1], where [NBBF1] is the concentration of the biomarker in non-blood fluids during the first time period, and [BB1] is the concentration of the biomarker in the blood during the first time period. Store the R ratio in memory; Biomarkers in non-blood fluids were measured during the second time period to determine [NBBF2], where [NBBF2] is the concentration of the biomarker in non-blood fluids during the second time period; and The measured [NBBF2] was correlated with the R ratio to generate the associated [BB2] biomarker concentration in the individual's blood during the second time period; It further includes periodically calibrating the ratio R to establish a recalibrated [NBBFX] / [BBX] ratio, where X is the nth time period, and n is 3 to 100; The calibration is selected from: a pair of calibration tests; or several pairs of calibration tests, and the calibration tests are performed for a period of time to establish the [NBBFA] / [BBA] average ratio, where [NBBFA] is the average biomarker concentration in non-blood fluids, and [BBA] is the average biomarker concentration in the blood of an individual during the time period.

2. The method of claim 1, wherein periodic calibration is performed over a time period selected from hourly, daily, weekly, monthly, semi-annual, annual, or combinations thereof, including intermediate values ​​and ranges.

3. The method of claim 1, wherein the establishment of the recalibrated [NBBFX] / [BBX] ratio is performed within at least one time period selected from: every minute, every hour, every morning, every noon, every night, every midnight, every day, every week, every month, every six months, every year, every two years, including intermediate time periods and ranges.

4. The method of claim 1, wherein the association is achieved by solving [BB2] in the following equation: [NBBF2] / [BB2]=R, Or rearrange and replace R: [BB2]=[BB1] / [NBBF1][NBBF2].

5. The method of claim 1, wherein the difference between the first time period and the second time period is at least one of the following: from 5 to 10 minutes, from 20 to 30 minutes, from 30 to 40 minutes, from 40 to 50 minutes, from 50 to 60 minutes, per hour, per day, per week, per month, per half year, per year, or per two years, including intermediate values ​​and ranges.

6. The method of claim 1, wherein storing the R ratio in the memory is implemented using a mobile communication device.

7. The method of claim 1, wherein the non-blood fluid is saliva, and the periodic calibration is performed at at least one interval selected from hourly, daily, weekly, monthly, semi-annual, annual, or combinations thereof, including intermediate values ​​and ranges.

8. The method of claim 1, further comprising applying machine learning (ML) to improve the accuracy of the method by manually comparing at least one of the following: preliminary results; secondary results; or tertiary results generated by a device having separated sample deposition regions and associated imaging devices or analyte concentration measurement devices.

9. A method for correlating glucose concentration in non-blood fluids with glucose in an individual's blood, comprising: Glucose in non-blood fluids and in the blood of the same individual were measured during the first time period to establish the [GNBF1] / [GB1] ratio (R), where [GNBF1] is the glucose concentration in non-blood fluids during the first time period and [GB1] is the glucose concentration in the individual's blood during the first time period. The [GNBF1] / [GB1] ratio is stored in memory; The glucose concentration in non-blood fluids was measured during the second time period, where [GNBF2] represents the glucose concentration in non-blood fluids during the second time period; and The measured [GNBF2] was correlated with the [GNBF1] / [GB1] ratio to generate the relevant estimated [BB2] glucose concentration in the individual's blood during the second time period; It further includes periodically calibrating the [GNBF1] / [GB1] ratio to establish a recalibrated [GNBFX] / [GBX] ratio, where X is the nth time period, and n is 3 to 100; The recalibrated [GNBFX] / [GBX] ratio is established within at least one of the following time periods: every minute, every hour, every morning, every noon, every night, every midnight, every day, every week, every month, every six months, every year, every two years, including intermediate time periods and ranges.

10. The method of claim 9, wherein the non-blood fluid is saliva.

11. The method of claim 9, wherein storing the [GNBF1] / [GB1] ratio in the memory is performed in a mobile communication device and analyzed by software in the mobile communication device.

Images